MulsCast

Applicability

Product	Supported
Atlas 350 Accelerator Card	√
Atlas A3 training product/Atlas A3 inference product	x
Atlas A2 training product/Atlas A2 inference product	x
Atlas 200I/500 A2 inference product	x
Atlas inference product AI Core	x
Atlas inference product Vector Core	x
Atlas training product	x

Function Usage

Multiplies src by scalar, converts the product to the half type in CAST_ROUND mode, and writes the result to dst based on mask. The formula is as follows:

$\text{[math]}$

Prototype

template <typename T0 = DefaultType, typename T1 = DefaultType, typename T2, RegLayout layout = RegLayout::ZERO, typename T3, typename T4>
__simd_callee__ inline void MulsCast(T3& dstReg, T4& srcReg, T2 scalarValue, MaskReg& mask)

Parameters

**Table 1** Template parameters
Parameter	Description
T0	Data type of the destination operand. For the Atlas 350 Accelerator Card, the supported data type is half.
T1	Data type of the source operand. For the Atlas 350 Accelerator Card, the supported data type is float.
T2	Data type of the scalar source operand. For the Atlas 350 Accelerator Card, the supported data type is float.
layout	RegLayout enum class. This API supports only RegLayout::ZERO and RegLayout::ONE. This parameter is used when the type of src is half. It does not take effect when the type is float. When the type is half, RegLayout::ZERO indicates that half elements are read from the even bits of b16 RegTensor and converted into float elements, and RegLayout::ONE indicates that half elements are read from the odd bits of b16 RegTensor and converted into float elements.
T3	RegTensor type of dstReg, for example, RegTensor<float>. It is automatically inferred by the compiler and does not need to be specified.
T4	RegTensor type of srcReg, for example, RegTensor<float>. It is automatically inferred by the compiler and does not need to be specified.

**Table 2** Parameters
Parameter	Input/Output	Description
dstReg	Output	Destination operand. The type is RegTensor.
srcReg	Input	Source operand. The type is RegTensor.
scalarValue	Input	Source operand. The type is scalar.
mask	Input	Valid indication of the source operand element operation. For details, see MaskReg.

Returns

None

Constraints

None

Example

template<typename T>
__simd_vf__ inline void MulsCastVF(__ubuf__ half* dstAddr, __ubuf__ float* srcAddr, float scalarValue, uint32_t count, uint32_t srcRepeatSize, uint32_t dstRepeatSize, uint16_t repeatTimes)
{
    AscendC::Reg::RegTensor<float> srcReg0;
    AscendC::Reg::RegTensor<half> dstReg0;
    AscendC::Reg::MaskReg mask;
    for (uint16_t i = 0; i < repeatTimes; i++) {
        mask = AscendC::Reg::UpdateMask<float>(count);
        AscendC::Reg::LoadAlign(srcReg0, srcAddr + i * srcRepeatSize);
        AscendC::Reg::MulsCast<half, float, float, AscendC::Reg::RegLayout::ZERO>(dstReg0, srcReg0, scalarValue, mask);
        AscendC::Reg::StoreAlign(dstAddr + i * dstRepeatSize, dstReg0, mask);
    }
}

Parent topic: Compound Computation